FlexiPulse: A machine-learning-enabled flexible pulse sensor for cardiovascular disease diagnostics
Zhiqiang Ma, Zhiqiang Ma, Haojun Hua, Changxin You, Zhihao Ma, Zhihao Ma, Guo Wang, Yang Xiao, Shirong Qiu, Ni Zhao, Yuan‐Ting Zhang, Derek Ho, Bryan P. Yan, Bee Luan Khoo
Abstract
Recently, the flexible pulse sensor has emerged as a promising candidate for real-time and population-wide monitoring of cardiovascular health. However, most current technologies are prohibitively expensive, lack clinical validation, or are not designed to diagnose cardiovascular disease (CVD) events. Here, we present the development of FlexiPulse, a low-cost, clinically validated, intelligent, flexible pulse detection system for CVD monitoring and diagnostics. The porous graphene-based FlexiPulse is prepared by eco-friendly and economical laser direct-engraving techniques and is feasible for mass production. FlexiPulse achieves high accuracy (>93%), as confirmed by clinical techniques, enabling it to precisely detect subtle changes in cardiovascular status. Furthermore, incorporating machine-learning algorithms in FlexiPulse allows it to perform independent clinical assessments of actual CVD events, including atrial fibrillation and atrial septal defect, with an average accuracy of 98.7%. We believe that FlexiPulse has the potential to promote remote monitoring and in-home care, thereby advancing precision medicine and personalized healthcare significantly.